dry forests and wildland fires of the inland northwest usa...
Post on 15-Jan-2020
8 Views
Preview:
TRANSCRIPT
Dry forests and wildland fires of the inland Northwest USA:
Contrasting the landscape ecology of the
pre-settlement and modern eras
Paul F. Hessburg a,*, James K. Agee b,1, Jerry F. Franklin b,2
a USDA Forest Service, Pacific Northwest Research Station, 1133 N. Western Avenue, Wenatchee, WA 98801-1229, USAb College of Forest Resources, P.O. Box 352100, University of Washington, Seattle, WA 98195-2100, USA
www.elsevier.com/locate/foreco
Forest Ecology and Management 211 (2005) 117–139
Abstract
Prior to Euro–American settlement, dry ponderosa pine and mixed conifer forests (hereafter, the ‘‘dry forests’’) of the Inland
Northwest were burned by frequent low- or mixed-severity fires. These mostly surface fires maintained low and variable tree
densities, light and patchy ground fuels, simplified forest structure, and favored fire-tolerant trees, such as ponderosa pine, and a
low and patchy cover of associated fire-tolerant shrubs and herbs.
Low- and mixed-severity fires provided other important feedbacks and effects to ponderosa pine-dominated stands and
landscapes. For example, in stands, frequent surface fires favored an ongoing yet piecemeal regeneration of fire-tolerant trees by
periodically exposing patches of mineral soil. They maintained fire-tolerant forest structures by elevating tree crown bases and
scorching or consuming many seedlings, saplings, and pole-sized trees. They cycled nutrients from branches and foliage to the
soil, where they could be used by other plants, and promoted the growth and development of low and patchy understory shrub
and herb vegetation. Finally, surface fires reduced the long-term threat of running crown fires by reducing the fuel bed and
metering out individual tree and group torching, and they reduced competition for site resources among surviving trees, shrubs,
and herbs. In landscapes, the patterns of dry forest structure and composition that resulted from frequent fires reinforced the
occurrence of low- or mixed-severity fires, because frequent burning spatially isolated conditions that supported high-severity
fires. These spatial patterns reduced the likelihood of severe fire behavior and effects at each episode of fire. Rarely, dry forest
landscapes were affected by more severe climate-driven events.
Extant dry forests no longer appear or function as they once did. Large landscapes are homogeneous in their composition and
structure, and the regional landscape is set up for severe, large fire and insect disturbance events. Among ecologists, there is also
a high degree of concern about how future dry forests will develop, if fires continue to be large and severe. In this paper, we
describe the key landscape pattern and process changes wrought by the sum of the settlement and management influences to
date, and we point to an uncertain future for ecosystem management. Widespread selection cutting of the largest and oldest
ponderosa pine and Douglas-fir in the 20th century has reduced much of the economic opportunity that might have been
associated with restoration, and long-term investment will likely be needed, if large-scale restoration activities are attempted. An
* Corresponding author. Tel.: +1 509 664 1722; fax: +1 509 664 1780.
E-mail addresses: phessburg@fs.fed.us (P.F. Hessburg), jagee@u.washington.edu (J.K. Agee), jff@u.washington.edu (J.F. Franklin).1 Tel.: +1 206 543 2688; fax: +1 206 543 3254.2 Tel.: +1 206 543 2138; fax: +1 206 543 3254.
0378-1127/$ – see front matter # 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.foreco.2005.02.016
P.F. Hessburg et al. / Forest Ecology and Management 211 (2005) 117–139118
uncertain future for ecosystem management is based on the lack of current and improbable future social consensus concerning
desired outcomes for public forestlands, the need for significant financial investment in ecosystem restoration, a lack of
integrated planning and decision tools, and mismatches between the existing planning process, Congressional appropriations,
and complex management and restoration problems.
# 2005 Elsevier B.V. All rights reserved.
Keywords: Pinus ponderosa; Abies grandis; Abies concolor; Pseudostuga menziesii; Landscape ecology; Mixed conifer forests; Fire ecology;
Fire history; European settlement; Historical range of variability
1. Introduction
Ponderosa pine (Pinus ponderosa) forests became
the poster child for unhealthy forests of the western
United States (US) around the time, when Gast et al.
(1991) presented their expert panel-based evaluation
of declining forest conditions in the Blue Mountains.
Thereafter, a number of additional qualitative and
quantitative assessments were undertaken to detail
changes that had occurred in the Inland West (Everett
et al., 1994; Hann et al., 1997; Hessburg et al., 1999a;
Huff et al., 1995; Lehmkuhl et al., 1994; O’Laughlin
et al., 1993; Perry et al., 1994; SNEP, 1996; USDA,
1993). But forest ecologists and fire scientists were
well aware of the importance of historical fires to these
environments decades before this time (e.g., see Arno,
1976, 1980; Arno and Allison-Bunnell, 2002; Cooper,
1961a,b; Daubenmire and Daubenmire, 1968; Gruell
et al., 1982; Hall, 1976; Morris, 1934a,b; Pyne, 1982;
Soeriaatmadja, 1966; Weaver, 1959, 1961).
Usually when people speak of ‘‘unhealthy’’ forests,
those prone to uncharacteristically intense or large-
scale fires, insect outbreaks, and epidemics of forest
diseases, they think of relatively dry ponderosa pine
forests that have experienced fire exclusion and heavy
selection cutting over several entries. The dry pine
forests of the Pinus ponderosa zone have certainly
been affected in this way, but dry mixed conifer
forests, primarily in the grand fir (Abies grandis),
white fir (Abies concolor), and Douglas-fir (Pseudot-
suga menziesii) zones (sensu Franklin and Dyrness,
1988), have also been similarly affected by 200 years
of settlement and management. We refer to forests in
these four zones collectively as the dry forest, and they
are represented by a range of cover types. Each cover
type in the dry forest was once dominated by
ponderosa pine cover under the influence of low-
and mixed-severity fire regimes of the last several
centuries. Without frequent low-severity or surface
fires, Douglas-fir, grand fir, white fir, and on the driest
sites ponderosa pine regenerated and released in the
understory, and have been growing there for more than
three-quarters of a century. To understand what makes
these forests currently ‘‘unhealthy’’, it is insightful to
examine the changes that have occurred to them over
the 200-year period of settlement and management.
2. Pre-settlement-era dry forests
Prior to the settlement of the West by European
immigrants, most dry forest environments were
burned by relatively frequent low and mixed-severity
fires (Arno and Allison-Bunnell, 2002). When we
refer to low-severity fires, we are describing fires that
occurred frequently, usually every 1–25 years, and
where less than 20% of the basal area was killed
(Agee, 1990, 1993). When we refer to mixed-severity
fires, we refer to fires that occurred with moderate
frequency, usually every 25–100 years, and where 20–
70% of the basal area may have been fire-killed. In the
context of dry forests, mixed-severity fires tended to
be at the lower end of this 20–70% overstory mortality
range. Such low- and mixed-severity fires favored
relatively low tree density and clumped tree distribu-
tion, light and patchy fuel beds, simple canopy
layering, and fire-tolerant tree and associated species
compositions (Fig. 1).
In stands, low-severity fires favored fire-tolerant
forest structures by removing the lower crown classes
(Fig. 2). These fires also cycled nutrients from foliage
and branches into the soil, promoted the growth of a
low and patchy shrub and herb cover, reduced the
threat of running crown fires by continually thinning
stands, eliminating fuel ladders, elevating crown
bases, and reduced competition for site resources
P.F. Hessburg et al. / Forest Ecology and Management 211 (2005) 117–139 119
Fig. 1. (A) Example of a typical single-layered, fire-tolerant, open ponderosa pine stand one would have seen under a low-severity fire regime in
the dry Pinus ponderosa zone (Franklin and Dyrness, 1988). Note the abundant bitterbrush (Purshia tridentata) shrub cover, the elevated tree
crown bases, the variably spaced trees, and the fire scars at the base of many trees. In stands, such as these, so-called ‘‘healthy’’ conditions would
have included some ongoing bark beetle mortality, especially in the tree clumps with higher stocking levels, and occasionally root disease
mortality. These processes contributed high-quality large snags and down wood. (B) Example of a ponderosa pine stand from the Pseudotsuga
menziesii zone. Note the advance regeneration of pine and Douglas-fir in the absence of recent fire.
Fig. 2. Example of a typical surface fire showing flame lengths and scorch heights less than 2 m. Above-ground portions of shrubs are scorched
and partially consumed, while root systems remain intact, thereby maintaining the edibility of many browse species.
P.F. Hessburg et al. / Forest Ecology and Management 211 (2005) 117–139120
among surviving trees, shrubs, and herbs (Kauffman,
1990). Frequent surface fires tended to favor the
largest trees with the thickest bark, producing an even-
aged appearance, but stands themselves exhibited
patchy age class distribution owing to the continuous
regeneration of trees by fires (Harrod et al., 1999).
In landscapes, the natural patterns of dry forest
structure and composition favored low- or mixed-
severity fires by maintaining a semi-predictable
mosaic, which spatially isolated conditions that
supported high-severity fires (Fig. 3). Hence, severe
fire behavior and fire effects were uncharacteristic of
dry forest-dominated landscapes (Hessburg and Agee,
2003; Hessburg et al., 1999a). Rarely, dry forest
landscapes were relatively more synchronized in their
vegetation and fuels conditions and affected by
climate-driven, high-severity fire events (Agee,
1997, 1998; Swetnam and Lynch, 1993; Whitlock
and Knox, 2002).
3. Modern-era dry forests
The most influential change agents of the settle-
ment and management periods can be grouped
according to their primary ecological effects. Change
agents acted to exclude fires, directly advance
secondary succession, suppress fires, or some combi-
nation of these effects.
Domestic livestock and wild ungulate grazing, road
and rail construction, grassland conversion to agri-
culture, urbanization, and rural development all
contributed to the direct or indirect exclusion of fires.
For example, domestic livestock grazing beginning in
the 1870s rather abruptly reduced the abundance and
distribution of flashy fuels that can rapidly carry
surface fires across the landscape (Belsky and
Blumenthal, 1997; Belsky et al., 1999; Irwin et al.,
1994). In addition, grazing by large populations of
deer during the 1940s–1960s, and more recently elk,
combined with domestic livestock grazing to remove
much of edible shrub cover as well as the flashy fuels
(Irwin et al., 1994; Riggs et al., 2000).
Road and rail construction fragmented broad forest
landscapes into smaller more isolated pieces, espe-
cially where road and rail beds functioned as effective
fuel breaks to potentially expanding low flame length
surface fires. Additionally, lands allotted to railroad
companies by Congress were sold to settlers or
harvested by the railroaders to pay the cost of building
the new rail lines. These changes in land allocation
and land use further fragmented the historical land-
scape with respect to historical fires (see Fig. 8 in
Hessburg and Agee, 2003).
Similarly, grassland conversion to agriculture
excluded fires, because many historical surface fires
in dry forests actually began on grassy benches, ridge
tops, or valley bottoms adjacent to dry forests and
woodlands, or in nearby shrub steppe communities,
and then migrated into dry forests. These fires were the
result of natural lightning ignitions or intentional
aboriginal burning (Agee, 1993, and references
therein; Barrett and Arno, 1982; Boyd, 1999, and
references therein; Robbins, 1999, and references
therein; Whitlock and Knox, 2002, and references
therein). Urbanization and rural development frag-
mented fire-prone dry forests at a fine scale by
substituting highly combustible vegetation types with
bare ground, concrete or asphalt, and grazed, irrigated,
or cultivated spaces.
Repeated selection cutting had the direct effect of
advancing secondary succession (Hessburg et al.,
1999a, 2000; Hessburg and Agee, 2003). The
preferred commercial species by land area and timber
volume affected were ponderosa pine and Douglas-fir,
and to a lesser extent western larch (Larix occidenta-
lis); these species were also important early seral
components that seeded in after fires (note that
Douglas-fir has broad ecological amplitude; it can
seed in after fire and tolerate shade). In addition, large-
diameter ponderosa pine and Douglas-fir were quite
tolerant of surface fires owing to their thick bark. As
these larger trees were removed by selective logging,
small canopy gaps were created and later filled by
small diameter Douglas-fir, grand fir, and white fir.
Fire exclusion and fire suppression also indirectly
contributed to advancing secondary succession by
preventing fire disturbances of a frequency and spatial
scale that would favor the dominance of the early seral
cover species.
Fire prevention and suppression still persist to this
day. While well intentioned, such suppression com-
pounds problems of advancing secondary succession
and the extreme fire intolerance and high contagion of
large expanses of dry forest. Small fires, if they had
been allowed to burn in the early 20th century, or were
P.F. Hessburg et al. / Forest Ecology and Management 211 (2005) 117–139 121
Fig. 3. Reconstructed historical (1900s) and current (1990s) maps of the Peavine Creek drainage, a dry forest subwatershed of the Lower Grand
Ronde subbasin in the Blue Mountains province displaying historical and current structural classes (A and B), fuel loading (C and D), crown fire
potential under average wildfire conditions (E and F), and flame length under average wildfire conditions (G and H), respectively (from Hessburg
and Agee, 2003). Structural class abbreviations are: si, stand initiation; se, stem exclusion (both open and closed canopy conditions); ur,
understory reinitiation; yfms, young multi-story forest; of, old multi-story and single story forest; nf, non-forest. Fuel loading classes are: very
low < 22.5 Mg/ha; low = 22.5–44.9 Mg/ha; moderate = 45–56.1 Mg/ha; high = 56.2–67.3 Mg/ha; very high > 67.3 Mg/ha. Crown fire poten-
tial classes were a relativized index. Flame length classes were: very low < 0.6 m; low = 0.7–1.2 m; moderate = 1.3–1.8 m; high = 1.9–2.4 m;
very high = 2.5–3.4 m; severe > 3.4 m.
P.F. Hessburg et al. / Forest Ecology and Management 211 (2005) 117–139122
intentionally lit, would have broken up the dry forest,
thereby reducing the size of the area influenced by
uncontrolled wildfires in the modern era. By virtue of
their political and social inertia, it is unlikely that
current fire prevention and suppression strategies will
change in a substantive way in the short term.
3.1. Ecologically important changes
What were some of the most important changes to
historical dry forest landscapes? The unintended chain
of settlement and management influences produced
ecologically unprecedented changes in dry forest
vegetation and its susceptibility to previously more
benign natural disturbances (Table 1). For example,
absence of the frequent surface fires and influenced by
grazing (Wissmar et al., 1994a), dry forests encroached
on nearby grasslands, shrublands, and meadows
(Fig. 4). Open stands of ponderosa pine or pine mixed
with Douglas-fir developed dense understories of
Douglas-fir, or Douglas-fir mixed with grand or white
fir (Fig. 5). Old forest patches were selectively logged or
clearcut leading to reduced old forest area and increased
fragmentation of the remaining patches (Fig. 6).
Except for the first quarter of the 20th century, most
wildfires were suppressed, when fires were smaller
Table 1
Key changes in the historical dry forest landscapes of the Inland West brou
et al., 2000; Hessburg and Agee, 2003)
Change
(1) Reduced grassland and shrubland area in forest potential
vegetation settings, and expanded woodland and forest area (Fig. 4)
(2) Reduced old and new forest area and connectivity (Fig. 6)
(3) Reduced area of fire-tolerant forest cover types, especially
ponderosa pine and western larch, and concordant increased
area of fire-intolerant forest cover types, especially Douglas-fir,
grand fir, and white fir (Fig. 7)*
(4) Loss of grass and shrub understories, and the concordant
addition of shade-tolerant conifer understories*
(5) Reduced forest area with large, early seral trees (Fig. 8)*
(6) Increased tree canopy cover, and canopy layers*
(7) Increased young multi-story forest area (Fig. 9)*
The asterisk (*) indicates a strong correlation with current severe fire be
than 4 ha (Agee, 1993, 1998). While seldom
discussed, this led to significantly reduced area and
connectivity of newly regenerated forest (i.e., early
seral shrub and grass areas in forest environmental
settings, and areas of new forest stand initiation
structure, Fig. 6, Hessburg et al., 1999a,c, 2000).
Fire-tolerant cover types, such as ponderosa pine
and western larch were harvested and widely replaced
by shade-tolerant Douglas-fir, grand fir, and white fir
cover types, most of which are also fire-intolerant due
to low crown bases, heat-trapping crowns, and thin
bark (Fig. 7). The fire-tolerant cover types were
harvested, because they contained the largest and
oldest high grade trees, hence the term ‘‘high grade’’
logging. Before settlement, remnant large trees
(>63.5 cm d.b.h.) of these same early seral species
were broadly distributed in other forest cover types,
and in forest structures that would not be considered
old forest, under most definitions (Hessburg et al.,
1999a, 2000). These scattered large trees occurred as a
lasting remnant after stand replacement disturbances
providing an important structural and habitat legacy
that would typically last for many centuries, first as a
standing live tree, then as a snag, and ultimately as
large down wood. These live remnant large trees were
widely eliminated by selection cutting (Fig. 8).
ght about by settlement and management influences (from Hessburg
Effect
(1) Increased homogeneity of the landscape vegetation
and fuels mosaic
(2) Increased homogeneity of the landscape vegetation
and fuels mosaic, reduced spatial isolation of areas
prone to high-severity fires
(3) Reduced fire tolerance, which favored: thin-barked
species, foliage that trapped more heat, and species
that typically have lower crown bases, thereby
maintaining continuity with fuel ladders
(4) Reduced likelihood of low-severity fires with increasing
flame length, fireline intensity, rate of spread, increased
fuel ladders and likelihood of crown fires
(5) Reduced fire tolerance (same as 3 above)
(6) Increased fuel ladders, potential flame lengths, fireline
intensity, rate of spread, and likelihood of crown fires
(7) Increased landscape homogeneity, reduced fire tolerance,
increased fuel ladders, potential flame lengths, fireline
intensity, rate of spread, and likelihood of crown fires
havior.
P.F. Hessburg et al. / Forest Ecology and Management 211 (2005) 117–139 123
3.2. Changes in disturbance processes
There were also many changes in disturbance
processes that were directly related to changes in
forest landscape attributes (Hessburg et al., 1999a,
2000; Hessburg and Agee, 2003) (Fig. 9). These
included: (1) elevated fuel loadings and increased
connectivity of high fuel loading; (2) increased
Fig. 4. Reconstructed historical (1900s) and current (1990s) maps of fores
Lower John Day subbasin of the Columbia Plateau province and (B) su
Mountains province (from Hessburg and Agee, 2003). Photograph (C) sho
meadow formerly maintained in grass and herb cover by frequent low-in
potential for running crowning fires; (3) increased
vulnerability to many insect and disease disturbances
of fire-intolerant tree species; (4) increased likelihood
of severe fire behavior in forest stands or patches with
respect to flame length, rate of spread, and fireline
intensity; (5) increased contagion or spatial aggrega-
tion of vulnerability to severe fire and insect and
disease disturbances (e.g., Figs. 3 and 10). Table 2
t and woodland cover type changes in (A) subwatershed 2701 in the
bwatershed 21 in the Lower Grande Ronde subbasin of the Blue
ws active encroachment of ponderosa pine (foreground) on a forest
tensity fires.
P.F. Hessburg et al. / Forest Ecology and Management 211 (2005) 117–139124
Fig. 4. (Continued ).
Table 2
Transitions from historical (ca. 1800) to current fire-severity in forested environments of the Interior Columbia River Basin (adapted from Hann
et al., 1997, and Quigley et al., 1996)
Historical to current transitions
Fire-severity class transition (historical to current) Percentage of historical forested area affected Hectares of historical forest affected
Lowa ! low 17.8 5,067,100
Low ! mixed 16.1 4,557,500
Low ! high 11.1 3,150,900
Subtotal (historical) 45.0 12,775,500
Mixed ! low 0.8 216,900
Mixed ! mixed 11.5 3,250,800
Mixed ! high 18.9 5,358,400
Subtotal (historical) 31.1 8,826,100
High ! low 2.4 681,900
High ! mixed 4.3 1,224,500
High ! high 17.2 4,879,900
Subtotal (historical) 23.9 6,786,300
Total 100.0 28,387,900
Historical and current fire-severity base maps are available at http://www.icbemp.gov.a Fire-severity classes used in Hann et al. (1997) were: ‘‘non-lethal’’, which is represented by ‘‘low’’, above; ‘‘mixed’’, which is represented by
‘‘mixed’’ and ‘‘lethal’’, which is represented by ‘‘high’’. Hann et al. (1997) defined lethal fire-severity as ‘‘stand replacement fires [that] leave less
than 20% of the basal area or less than 10% of the overstory vegetation [cover] that was living prior to the fire.’’ The latter clause applied to lethal
fires of shrub and herb physiognomies that are not reported here. Hann et al. (1997) defined non-lethal fire-severity as ‘‘fires [that] leave more than
70% of the basal area or more than 90% of the canopy cover that was living prior to the fire.’’ The latter clause applied to non-lethal fires of shrub and
herb physiognomies that are not reported here. They defined mixed fire-severity to ‘‘include all fires of intermediate effects’’, that is, intermediate to
lethal and non-lethal fire-severity. Intermediate effects would come from fires having both lethal and non-lethal components, and fires would leave
20–70% of the basal area or 10–90% of the overstory vegetation cover. These definitions are consistent with those used in this paper.
P.F. Hessburg et al. / Forest Ecology and Management 211 (2005) 117–139 125
Fig. 5. Example of how an open ponderosa pine stand appears after heavy selection cutting and decades of in-growth by Douglas-fir (A). White
arrows show two remaining large ponderosa pine after selection cutting. From above (B), one can readily see the filling in of the landscape with
shade-tolerant, pole- to medium-sized trees.
P.F. Hessburg et al. / Forest Ecology and Management 211 (2005) 117–139126
Fig. 6. Reconstructed historical (1900s) and current (1990s) maps of the Libby Creek drainage (Methow 11), a dry forest subwatershed of
Methow subbasin in the Northern Glaciated Mountains province displaying structural classes (from Hessburg et al., 1999a). Structural classes
are: si, stand initiation; seoc, stem exclusion, open canopy; secc, stem exclusion, closed canopy; ur, understory reinitiation; yfms, young multi-
story forest; ofms, old forest, multi-story, and non-forest.
shows the transitions in fire-severity from historical
(ca. 1800) to existing conditions in the Interior
Columbia River Basin (adapted from Hann et al.,
1997, Quigley et al., 1996). Table 3 summarizes the
change in hectares and change from historical
percentage area of the Interior Columbia Basin by
fire-severity class. By examining both tables, one may
quickly see that in the Inland Northwest, the scale of
fire regime shifting has been enormous. There has been
a net reduction of low-severity regime area in the Basin
of 53%, the area of mixed-severity fire is roughly the
same (although the area has moved from characteristic
to uncharacteristic sites, Table 2), and the area of high-
severity fire has nearly doubled (Table 3).
Surface fuels became elevated primarily in two
ways: fuels created by repeated selection cutting were
often left untreated, and stands that resulted from
selection cutting became densely stocked, layered,
Table 3
Change in actual and percentage area of fire-severity classes from the histori
Basin (adapted from Hann et al., 1997, and Quigley et al., 1996)
Fire-severity class Historical area (ha) Current area (ha)
Lowa 12,775,500 5,965,900
Mixed 8,826,100 9,032,800
High 6,786,300 13,389,200
Total 28,387,900 28,387,900
Historical and current fire-severity base maps are available at http://wwwa See footnote (a) in Table 2.
and composed of tree species that were vulnerable to a
broad array of tree-killing insects and pathogens. The
dense stocking and layering was due to the ongoing
regeneration and release of shade-tolerant trees in the
gaps left by each successive cutting, and it created
highly effective fuel ladders. Decades of ongoing
forest insect and disease mortality also contributed to
present-day fuel loads. High connectivity of expanded
fuel loads occurred, because frequently surface fired
historical ponderosa pine forests tended to be spatially
well connected. Repeated selection cutting of these
same patterns left well-connected fuel beds.
The likelihood of running crown fires increased,
because selection cutting and fire exclusion favored
increased tree cover, canopy layering, and tree density.
The increased tree cover consisted of shade-tolerant
trees with low crown bases and heat-trapping foliage.
The increase in canopy layers ensured adequate fuel
cal (ca. 1800) to the existing condition in the Interior Columbia River
Change in area (ha) Percentage of historical area
�6,809,600 47
206,700 102
6,602,900 197
.icbemp.gov.
P.F. Hessburg et al. / Forest Ecology and Management 211 (2005) 117–139 127
Fig. 7. Reconstructed historical (1900s) and current (1990s) maps of forest cover types in: (A) Peavine Creek drainage, a dry forest subwatershed
of the Lower Grand Ronde subbasin in the Blue Mountains province and (B) subwatershed 09 in the Pend Oreille subbasin of the Northern
Glaciated Mountains province (from Hessburg et al., 1999a). Cover type abbreviations are: PIPO, ponderosa pine; LAOC, western larch; PICO,
lodgepole pine; PSME, Douglas-fir; ABGR, grand fir; ABLA2/PIEN, subalpine fir and Engelmann spruce; TSHE/THPL, western hemlock and
western redcedar; PIMO, western white pine; HDWD, hardwoods.
ladders for conveying fire from the forest floor to the
overstory canopy. Similarly, fireline intensity, fire rate
of spread, and flame length attributes increased in
severity because of the expanded coverage of highly
combustible surface fuels, and the increased layering
and spatial extent of understory fire-intolerant trees
(Huff et al., 1995).
4. Implications for future management
Much of the increased area in present-day mixed
and high-severity fire regimes developed in the dry
forests of the Douglas-fir, grand fir, and white fir
zones, where ponderosa pine was the primary early
seral species. As an alternative to blanket prescriptions
P.F. Hessburg et al. / Forest Ecology and Management 211 (2005) 117–139128
Fig. 8. Reconstructed historical (1900s) and current (1990s) maps showing forest patches with remnant large trees in Lower Yakima 100, a dry
forest subwatershed of the Lower Yakima subbasin in the Northern Cascade Mountains province (from Hessburg et al., 1999a).
that simply reduce fuels and thin dry forests, we
suggest a more ecological approach is warranted; one
that would restore more natural spatial and temporal
patterns of dry forest structure, composition, snags,
and down wood (e.g., Hessburg et al., 1999b,c). It was
the broad natural (including aboriginal inputs) range
of spatial patterns and temporal variation in those
patterns of living and dead forest, grassland, shrub-
land, and woodland structure that historically sup-
ported the native species and processes that were
inherited at the time of settlement (Lindenmayer and
Franklin, 2003; Thompson and Harestad, 2004; Turner
and Romme, 1994).
The suggested approach would involve reconstruct-
ing, using empirical, simulation, or some combination
of methods, a representative portrayal of the broad
natural variation in vegetation patterns that existed
prior to management and settlement (e.g., see
Hessburg et al., 2004; Keane et al., 2002; Reynolds
and Hessburg, 2004). Landscape management of the
dry forests based on the knowledge of natural patterns
of structure and composition would inevitably lead to
more characteristic fire regimes, and widespread
reductions in dead wood in the form of snags and
down wood, as levels of these components are
currently elevated in present-day dry forests. These
more ecologically motivated management actions
would enable the restoration of vegetation pattern and
disturbance process interactions that support the fire
regimes, forest resources and values, and other
ecosystem processes that society is apparently
interested in (H.R., 1904, U.S. Government, 2003).
An evaluation of changes in fire regimes of the
Interior Columbia River basin and the Central Sierra
Nevada ecosystems over the last 200 years reveals that
many dry forest landscapes will be burned by mixed or
high-severity fires in the coming years (Hann et al.,
1997; SNEP, 1996). At watershed to subregional
scales, most dry forest landscapes in their current
condition are set up for large, severe fire events (Figs. 3
and 11). It is apparent that the choice available to
decision makers and citizens alike is the type of fire and
smoke they will choose—that associated with wildfires
or that which is actively prescribed and managed.
The effects of wildland fires of the last 10–15 years
alone suggest that some sort of active management is
needed to increase the likelihood that managers can
produce the desired amount, timing, and spatial
arrangement of treatment effects on current dry forest
landscapes. To date, wildland fires alone have not
created ecological outcomes that are desired by society
or that are consistent with natural ecosystem function-
ing. Even with active management, there is no
guarantee that managers will always achieve desired
results, because environments and ecosystem structures
and processes all have strongly associated stochastic
P.F. Hessburg et al. / Forest Ecology and Management 211 (2005) 117–139 129
Fig. 9. Reconstructed historical (1900s) and current (1990s) maps of: (A) Lower Grande Ronde 21 and (B) the Peavine Creek drainage (Lower
Grande Ronde 55), dry forest subwatersheds of the Grande Ronde subbasin in the Blue Mountains province displaying structural classes (from
Hessburg et al., 1999a). Structural classes are: si, stand initiation; seoc, stem exclusion, open canopy; secc, stem exclusion, closed canopy; ur,
understory reinitiation; yfms, young multi-story forest; ofms, old forest, multi-story; ofss, old forest, single story, and non-forest.
features (Lewontin, 1966; Reckhow, 1994; Walters and
Hollings, 1990).
It is likely that various combinations of pre-
commercial and commercial thinning, mechanical
slash treatments, and prescribed burning will be used
at a stand-level to produce the desired effects on fuels
and vegetation, because that is what managers are
skilled at doing. Furthermore, several sequential
applications of prescribed burning may be needed,
because the first burn treatments may temporarily
increase surface fuels (Agee et al., 2000; Agee, 2002).
At a landscape-level, recent fires and those of the
last several centuries have taught us that managers
may have more success in planning and managing for
P.F. Hessburg et al. / Forest Ecology and Management 211 (2005) 117–139130
Fig. 10. Reconstructed historical (1900s) and current (1990s) maps of: (A) Kettle 20 and (B) Wallowa 29, dry forest subwatersheds of the Kettle
and Wallowa subbasins, in the Northern Glaciated Mountains and Blue Mountains provinces, respectively. Maps show changes in vulnerability to
western spruce budworm (A) and Douglas-fir dwarf mistletoe (B). Changes in vulnerability are related to changes in vegetation structure and
composition (see Section 2 for estimating vulnerabilities in Hessburg et al., 1999a,d).
dynamic landscapes rather than predetermined static
or reserved conditions. Inability to control chance
events, climatic regime shifting, and the stochastic
features of environments and ecosystem processes
compel one to manage for desired future dynamics
rather than desired future conditions. That is, one will
likely be more successful managing forests within a
range of conditions, rather than in specific states.
5. Planning priorities for altering fuels and fire
behavior
When considering treatment priorities in restora-
tion planning, several simple guidelines emerge. The
stylized concept of historical fire regime areas can be
used as a starting basis or template (Cissel et al., 1999;
Agee, 2003), with local alterations for overlapping
P.F. Hessburg et al. / Forest Ecology and Management 211 (2005) 117–139 131
Fig. 11. Broadscale (1-km2 pixels) map of current fire-severity classes with transitions from historical in the Interior Columbia Basin and
vicinity (adapted from Hann et al., 1997). High severity = stand-replacing fire that kills > 70% of the overstory tree basal area; low severity = fire
that kills < 20% of the overstory tree basal area; mixed severity = fire that kills 20–70% of the overstory tree basal area.
regime areas as is appropriate. We describe current
priorities for restoration planning in terms of the land
area of the historical fire regime, because it is spatially
the most accurate and succinct way of describing the
actual places that could most benefit from restoration.
The problem is that the current area of the high and
mixed-severity fire regimes is a mixture of places that
have changed very little, in terms of fire regime
parameters, and of places that have changed a great
deal. Tables 2 and 3 illustrate this point.
5.1. Historical low-severity regime areas
Land areas that historically supported low-severity
surface fires are, from an ecological perspective, the
highest priority for fuel reduction and vegetation
pattern restoration treatments (Hessburg and Agee,
2003). These are the landscapes dominated by the dry
forests. The structure, composition, and spatial patterns
of historical dry forest landscapes tolerated fires well,
and these conditions and their variability can be used as
a basis for restoring more characteristic fires. The
principles of fire-tolerant or fire-resilient (‘‘FireSafe’’)
forests of Agee and others (Agee, 2002; Agee et al.,
2000; Hessburg and Agee, 2003) can be applied
through prescribed burning and with the use of
silviculture at the stand-level, and these are relatively
well-accepted principles. But as one scales up from
individual patches to landscapes, the priority of where
to treat and how much to treat becomes much less
clear, and it will be essential to define spatial and
temporal patterns in upcoming management decisions.
One of the challenges faced by research and
management is to quantify the spatial patterns, temporal
P.F. Hessburg et al. / Forest Ecology and Management 211 (2005) 117–139132
variation, and amounts of vegetation and fuel treat-
ments that are needed to make current landscapes more
fire-resilient and at the same time supportive of native
species, their habitats, and related processes through
time. Modeling efforts are underway to begin to address
the fire resilience component of this need (e.g., see
Finney, 2001), but practical landscape experiments are
also needed on the ground to evaluate and authenticate
model outputs, and understand controlling factors and
variation. Reference conditions provide useful insights
into the ranges of vegetation and fuels spatial patterns
that have supported low-severity fire regimes and native
species habitats in the past (Agee, 2003; Fule et al.,
1997; Hessburg et al., 1999b,c; Landres et al., 1999;
Swetnam et al., 1999).
5.2. Historical mixed-severity regime areas
Also from an ecological perspective, land areas that
historically supported mixed-severity fires are the
second priority for fuel reduction and vegetation
pattern restoration treatments (Hessburg and Agee,
2003). These are the landscapes dominated by both
dry and mesic mixed conifer forests. In the past, the
importance of these mixed-severity landscapes has
been understated, because managers and scientists,
trying to convey a simple and concise story about
change in dry forests, focused their attention on forests
that were burned by frequent surface fires. But it is
increasingly clear that throughout the Inland North-
west, many mixed conifer forests were burned by both
low- and mixed-severity fires (Agee, 1993, 1998;
Hessburg and Agee, 2003). This variation in fire
regime provided additional variation in resulting
patterns of living and dead vegetation, habitats, and
related ecological processes.
At the present time, the least is known about
restoring patterns and functionality to forests of the
historical mixed-severity regime, because mixed-
severity fires represent the broadest range of fire
effects. Reasonable hypotheses can be derived from
reconstructions of historical landscapes about how
much area to treat, what kind of mosaic to leave, and
which treatments may be functional fire surrogates,
but they should be tested. Again, estimates of
historical range and variation in vegetation and fuels
patterns can provide valuable insights into the spatial
patterns and temporal variation in patterns that
supported the full range of mixed-severity fires and
their effects.
6. Landscape restoration considerations
6.1. Buying time for forest restoration
It has taken more than a century for the conditions
to develop that exist today, and at best, it may take
several decades to make substantive and widespread
improvements. Currently, dry forest landscapes of the
Inland Northwest exhibit high landscape connectivity
of conditions that support large and severe fires. To
buy time for more thoughtful and carefully planned
forest restoration, it makes sense to begin restoration
by designing and developing networks of shaded
fuelbreaks throughout the dry forests (Agee et al.,
2000; Arno and Allison-Bunnell, 2002). These net-
works would provide the advantage of breaking large
fire-prone landscapes into smaller and more manage-
able pieces, which would be of significant benefit,
both for restoration and fire suppression efforts. It
would be useful to position fuelbreaks adjacent to
existing roads so that the fuelbreaks could be revisited
at regular intervals, and re-treated to maintain a widely
scattered cover of medium and large-sized ponderosa
pine cover (where available) with only light fuels.
Fuelbreaks could also make wise use of topography
so that fire managers could use them as boundaries to
implement broader fuel treatments, and fire fighters
could use them to burn up against in fire suppression.
For example, ridgetop topographies and valley
bottoms adjacent to major stream courses and stream
confluences (but well outside of the riparian zone)
may be suitable locations (Agee et al., 2000; Camp
et al., 1997). The intent with shaded fuelbreaks is to
allow fire to burn through the fuelbreak area, but to
constrain fire behavior to that of a low-intensity
surface fire (Agee et al., 2000; Anderson and Brown,
1988; Schmidt and Wakimoto, 1988).
Many fuelbreaks can be created by thinning. To be
effective, it is postulated that fuelbreaks should be at
least 400-m wide to improve chances of inhibiting the
spread of a running crown fire (Agee et al., 2000;
Anderson and Brown, 1988; Finney, 2001). Wildfires
can move through fuelbreaks, but usually as surface
fires, except when fire weather and fire behavior are
P.F. Hessburg et al. / Forest Ecology and Management 211 (2005) 117–139 133
extreme. In these cases, even broad fuelbreaks can be
jumped for great distances by blowing fire brands and
spotting fire. Thinning should remove the smaller trees
which make up the fuel ladders. Crown fire hazard can
be reduced by removing enough of the taller trees to
reduce forest crown cover to less than 35% and maintain
a minimum open space between the crowns of the taller
trees of at least 10 feet (Schmidt and Wakimoto, 1988).
However, such removal can also encourage regenera-
tion of shrubs and trees, such that the required
maintenance interval shortens significantly. While
removal of surface fuels and fuel ladders is important,
treatment of the overstory involves more tradeoffs and
may notalwaysbepart ofa successful treatment strategy
(Agee and Skinner, this volume). Post-treatment fuels
may be piled and burned, or broadcast burned, where
feasible to increase treatment efficacy, and fuelbreaks
may be retreated often enough that only a low intensity
surface fire could be supported by the fuelbed.
6.2. Improving fire resilience with silviculture and
prescribed burning
Added to fuelbreaks, the fire resilience of dry forest
landscapes can be improved by thinning from below,
free thinning, applying shelterwood regeneration
harvest with reserves, and by coupling prescribed
burning treatments with silvicultural treatments over
large areas (Agee et al., 2000; Graham et al., 1999,
2004). Initially, small drainages or subwatersheds
could be considered for applying area-wide treatments
(Arno and Allison-Bunnell, 2002). Usually, fuelbed
and tree density conditions have changed the most on
the southerly dry aspects, and these can be the first
areas designated for treatment. By treating southerly
aspects initially, large landscapes composed of
contagious fuelbeds and dense and layered tree cover
can be broken up in a highly intuitive manner.
The effectiveness of treatments, whether thinning
or shelterwood regeneration harvests, may be
enhanced by emphasizing opening up the canopy to
relatively wide spacing, reducing canopy layering,
removing of the most fire-intolerant species, usually
grand and white fir, but also Douglas-fir, and removing
the smaller size classes. These are the ordinary effects
of surface fires. Arno and Allison-Bunnell (2002)
suggest that historical surface fire regimes perpetuated
ponderosa pine-dominated stands with 30–100 trees
per acre. Harrod et al. (1999) suggest similar wide
spacing. Trees that are left behind may typically be
those that are the most vigorous, are largest, and have
the thickest bark. Generally, ponderosa pine may be
favored over other species in the mix of leave trees, but
Douglas-fir, another natural component, can be left as
well, especially where trees are large and thickly
barked, vigorous, and well-spaced. In general,
silvicultural systems that are applied to manage dry
forest stands can be designed to reduce the potential
for a running crown fire. Selection thinning and crown
thinning (thinning from above) that maintain multiple
layers typically may not reduce the risk of crown fires
and tree torching nearly as much as low thinning, free
thinning, and shelterwood harvests with reserves
(Graham et al., 1999).
Obviously, if only a few of the best methods were
applied everywhere to improve the fire resilience of
Inland Northwest dry forests, the structure and
composition of dry forests could become overly
simplified, and biotic diversity and ecosystem function-
ing could be adversely impacted. No single thinning or
management prescription, therefore, may be expected
to achieve all objectives across any given landscape. By
understanding the stand and landscape characteristics
that affect prescribed or wildfire behavior, and that are
necessary to support native species and processes,
forest landscapes may be designed through the use of
silviculture and prescribed burning that minimize the
potential for severe fire effects or crown fire behavior.
6.3. Using historical spatio-temporal variation
as a basis for laying out treatments
To a large extent, what drives uncharacteristic fires
is pronounced change in landscape patterns of fuels,
forest structure, and composition (Hessburg and Agee,
2003) coupled with ordinarily high potential for severe
fire weather in the West (Agee, 1997, 1998; Franklin
and Agee, 2003). Over long-time frames, fires, insect
outbreaks, disease epidemics, and weather events
historically created and maintained patterns of dry
forest structure and composition that supported an
exceptional variety of plant and animal species, and a
host of critical processes. The interplay between
patterns and processes created a metastable patch
dynamic (Ahl and Allen, 1996; Allen and Starr, 1982;
Wu and Loucks, 1995), a broad set of conditions that
P.F. Hessburg et al. / Forest Ecology and Management 211 (2005) 117–139134
strongly tended to support surface fires over other
kinds. Native species and processes are at best adapted
to and at worst showed they could persist through
these patch dynamics, and perhaps more importantly,
native species and processes showed that they could
persist within the context of the rates of change that
were characteristic of dry forest ecosystems.
Characteristic patterns of forest structure and
composition have been replaced by new and untested
sets of patterns that resulted from an unintended chain
of influences. Furthermore, rates of ecosystem and
habitat change in the last two centuries appear to be
much more rapid than was formerly the case. This may
be particularly important, when considering native
species viabilities in the long term, and it may explain
why there are so many native plant and animal species
currently listed as sensitive, threatened, or endangered.
Current patterns of dry forest landscapes primarily
support mixed and high-severity fires, and these fires
are occurring over much broader landscapes than was
formerly the case. For example, patches of isolated
stand replacement fire were common in historical dry
forest landscapes, but today, entire landscapes are
claimed by severe fires (Fig. 3). Furthermore, present-
day large wildfires synchronize landscapes by creating
very large patches with corresponding forest regen-
eration, species composition, structure, fuel beds, and
size and age class distribution, thereby facilitating
very large future wildfires.
To create fire regimes that are more predictable and
more consistent with environmental settings under the
current climatic regime, we suggest that landscape
patterns of fuel, forest structure, and composition will
need to be created that are characteristically associated
with those regimes. We further suggest, that to
improve assurances that native species and processes
will persist, it will also be important to restore forest
landscapes that reflect some semblance of the spatial
and temporal variation in patterns that species evolved
with. Dramatic and rapid departures in fire regimes
and landscape patterns portend losses to species and
uncharacteristic changes to vital processes.
6.4. Dealing with the past while heading into
the future
Many dry forests may be in dire need of pattern
restoration, but there is a legacy associated with past
management actions, and restorative management
treatments may not be possible in the short-term
everywhere they may be needed. For example, the
current code of federal laws requires that remaining
strongholds for native anadromous and cold water
fishes listed under the Endangered Species Act must
be maintained as landscape patterns are restored (Lee
et al., 1997). Hence, when considering both listed fish
and landscape restoration needs, it may be important
to prioritize and sequence landscape restoration needs
at continental, regional, and subregional scales, and to
consider a corresponding watershed hierarchy (see
Rieman et al., 2000).
Similarly, timber harvests have so reduced the
natural stores of large trees that it seems important to
conserve many of the remaining populations for their
biological and ecological values alone (Lindenmayer
and Franklin, 2003; Franklin et al., 1981). Character-
istic late-successional and old forests are in very short
supply as is landscape area with remnant large trees
(Hann et al., 1997; Hessburg et al., 1999a; SNEP,
1996). If motivated for ecological restoration, future
thinning and prescribed burning, therefore, could
primarily target removal of the small and medium-
sized fire-intolerant trees. Such ecologically motivated
restoration in the Inland Northwest could focus on
managing for more and better connected areas of
characteristic late-successional and old forests and
areas with remnant large trees, and adding back these
elements in characteristic patterns.
Large-scale restoration will be costly in many
places. Widespread selection cutting of the largest and
oldest ponderosa pine and Douglas-fir has reduced the
economic opportunity that might have otherwise been
associated with dry forest restoration. Large-scale and
long-term investment in restoration treatments will be
needed in some landscapes to return a semblance of
characteristic patterns and processes.
7. Conclusions
Dry forests of the present-day no longer appear or
function as they once did. Current patterns of forest
structure and composition do not resemble even recent
historical conditions, neither do they represent what
we would expect to see under or more natural or
characteristic disturbance regimes and the current
P.F. Hessburg et al. / Forest Ecology and Management 211 (2005) 117–139 135
climate. There is little evidence that current patterns
are sustainable and this has important ecological
consequences.
Large landscapes are increasingly homogeneous in
their composition and structure, and the regional
landscape is set up for severe, large fire and insect
disturbance events. Among ecologists, there is a high
degree of concern about how future dry forests will
developandwhat theywillbecome, iffirescontinuetobe
large and severe. Restoration of forest patterns with high
ecological functionality will not be easy and solutions
will not be clear cut. Furthermore, the possibility that
certain changes in ecosystems may be chronic (i.e.,
irreversible, from a practical standpoint) may make full
restoration of patterns and processes impossible.
7.1. What can be done?
Much can be done to make progress in restoring
pattern and process of dry forest ecosystems. Managers
may make great headway in modifying the structure
and composition of dry forests using various combina-
tions of commercial and pre-commercial thinning and
prescribed burning treatments. They can favor fire-
tolerant species like ponderosa pine and western larch
in their treatments and discriminate against fire-
intolerant species, such as Douglas-fir, grand fir, and
white fir. They can favor leaving the large and very
large trees, and remove the small and medium-sized
trees. Managers can treat large areas, but first they must
be mindful of the consequences of the past and shape
their future courses of action with that understanding.
These considerations can influence, where they treat,
how they treat, and when they will treat.
Managers can study the natural vegetation and fire
patterns of areas they intend to manage and other areas
like it, and apply their knowledge of those patterns to
their management (Agee, 2003; Hessburg et al.,
1999b,c). And they can continue preventing and
suppressing fires, at least until more characteristic
patterns of vegetation and fire regime are restored.
Finally, managers may need to apply relatively
frequent follow up treatments to already treated stands
and landscapes to maintain their fire-tolerance. Ideally,
to control costs and provide desirable fire effects, this
would include a program of regular underburning.
The current planning environment for forest
management in the Inland West is exceedingly complex
in terms of both issues and information available for
issue resolution. From recent bioregional assessment
projects (e.g., Columbia River Basin and Sierra Nevada
Ecosystem Projects), we now know more about existing
forest and rangeland conditions, and the state of
ecosystems and their inhabitants than ever before.
Restoration and management planning requires ade-
quate incorporation and integration of this information
to address interactions between landscape spatial and
temporal patterns of vegetation, habitats, fuels, and
potential fire behavior. Planning further should consider
the relations between these conditions and a host of
issues surrounding terrestrial habitats and associated
aquatic ecosystem conditions. Before management
alternatives can be selected and implemented, they
must be adequately evaluated for their effects on many
terrestrial and aquatic ecosystem components.
In light of the above planning requisites, we believe
there is a critical need for modeling and simulation
tools that not only incorporate multidimensional data
and examine the effects of management activities on
key facets of ecosystems, but that also serve as
platforms for integrating and evaluating the simulta-
neous effects of alternatives on a host of ecosystem
conditions. While many modeling and simulation,
scenario planning and decision support tools have
been developed in relative isolation in the last decade
(see discussion in Hessburg and Agee, 2003), their use
in planning is poorly established. Increased emphasis
and investment in the further development, acceptance
and application of these tools in the planning
environment would seem extremely important in the
future.
A major challenge for restoring pattern and process
in Interior Northwest forests relates to the complex and
sometimes contentious sociopolitical context within
which forest management is planned and implemen-
ted. It may be unlikely that effective long-term
management options can be developed through social
consensus due to variation in values among different
sectors of society, fluctuating social values over time,
and a tendency of consensus-based options to favor
little change over the status quo (McCool et al., 1997).
From a public policy standpoint, however, the recently
enacted public law H.R., 1904, also known as the
Healthy Forests Restoration Act (HFRA, U.S. Govern-
ment, 2003) has emerged as a major influence on
future forest restoration planning and implementation.
P.F. Hessburg et al. / Forest Ecology and Management 211 (2005) 117–139136
HFRA will result in vastly expanded efforts by land
managers in coming years in support of fuel reduction
and forest-thinning projects within the dry forests of
the Inland West, concurrent with limited judicial
review. Recognizing that re-establishment of healthy
forests is a dynamic, long-term (decades to centuries)
process (Hessburg and Agee, 2003), we believe that
the effectiveness of HFRA activities may be greatly
enhanced, if due attention is given to ongoing
monitoring of both the ecological effects of treat-
ments/projects, and their effectiveness in meeting
prescribed objectives. In further recognition that
restoration is a long-term investment, addressing
and emphasizing the needs and long-term financial
support for repeated (over time) thinning, fuel
reduction and other vegetation management efforts
would also seem extremely important in HFRA
implementation. Lastly, we believe that consideration
of two critically important premises could greatly
motivate and improve the implementation of HFRA:
(1) an upfront admission that neither scientists nor
managers have all, or even many, of the answers to
questions of how to restore more healthy forest
conditions and (2) acknowledgement that once
restored, dry forests should not only support the fire
regime of interest, but also viable populations of native
species in functional habitat networks across space
and through time. If embraced, the first premise would
call for an experimental, adaptive approach to
management (sensu Bormann et al., 1999; Franklin,
1993; Thomas and Huke, 1996), operationally uniting
research and management in planning, implementa-
tion and monitoring of treatments. The second premise
would call for formulating fuels reduction, thinning
and other vegetation treatments in the context of
improved landscape functionality for species and vital
ecosystem processes.
Acknowledgments
This research was supported in part by the
Managing Disturbances Program of the Pacific
Northwest Research Station (RWU-4577), USDA
Forest Service. Ed Depuit and three anonymous
reviewers provided helpful reviews of an earlier draft.
We thank Brion Salter for his excellent assistance in
developing the figures and tables.
References
Agee, J.K., 1990. The historical role of fire in Pacific Northwest
forests. In: Walstad, J.D., Radosevich, S.R., Sandberg, D.V.
(Eds.), Natural and Prescribed Fire in Pacific Northwest Forests.
Oregon State University Press, Corvallis, OR, 317 p.
Agee, J.K., 1993. Fire Ecology of Pacific Northwest Forests. Island
Press, Washington, DC, 493 pp.
Agee, J.K., 1997. The severe weather wildfire: too hot to handle?
Northwest Sci. 71, 153–157.
Agee, J.K., 1998. The landscape ecology of western forest fire
regimes. Northwest Sci. 72, 24–34.
Agee, J.K., 2002. The fallacy of passive management. Conserv. Biol.
Pract. 3 (1), 18–25.
Agee, J.K., 2003. Historical range of variability in eastern Cascades
forests, Washington, USA. Landscape Ecol. 18, 725–740.
Agee, J.K., Bahro, B., Finney, M.A., Omi, P.N., Sapsis, D.B., van
Wagtendonk, J.W., Skinner, C.N., Weatherspoon, C.P., 2000.
The use of shaded fuelbreaks in landscape fire management.
Forest Ecol. Manage. 127, 55–66.
Agee, J.K., Skinner, C.N. Basic principles of forest fuel reduction
treatments. Forest Ecol. Manage., this volume.
Ahl, V., Allen, T.F.H., 1996. Hierarchy Theory: A Vision, Vocabu-
lary, and Epistemology. Columbia University Press, USA, New
York, NY.
Allen, T.F.H., Starr, T.B., 1982. Hierarchy: Perspectives for Ecolo-
gical Complexity. University of Chicago Press, USA, Chicago,
IL.
Anderson, H.E., Brown, J.K., 1988. Fuel characteristics and fire
behavior considerations in the wildlands. In: Fischer, W.C.,
Arno, S.F. (Eds.), Protecting People and Homes from Wildfire
in the Interior West. Proceedings from the Symposium and
Workshop. General Technical Report INT-GTR-251. U.S.
Department of Agriculture, Forest Service, Intermountain
Research Station, Ogden, UT, pp. 124–130.
Arno, S.F., 1976. The historical role of fire on the Bitterroot National
Forest. Research Patent INT-187, U.S. Department of Agricul-
ture, Forest Service, Intermountain Forest and Range Experi-
ment Station, Missoula, MT, 29 pp.
Arno, S.F., 1980. Forest fire history in the Northern Rockies. J.
Forestry 78, 460–465.
Arno, S.F., Allison-Bunnell, S., 2002. Flames in Our Forest: Disaster
or Renewal. Island Press, Washington, DC, 227 p..
Barrett, S.W., Arno, S.F., 1982. Indian fires as an ecological
influence in the Northern Rockies. J. Forestry 80, 647–651.
Belsky, J.A., Blumenthal, D.M., 1997. Effects of livestock grazing
on stand dynamics and soils in upland forests of the interior west.
Conserv. Biol. 11 (3), 315–327.
Belsky, J.A., Matzke, A., Uselman, S., 1999. Survey of livestock
influences on stream and riparian ecosystems in the western
United States. J. Soil Water Conserv. 54, 419–431.
Bormann, B.T., Martin, J.R., Wagner, F.H., Wood, G., Alegria, J.,
Cunningham, P.G., Brookes, M.H., Friesma, P., Berg, J., Hen-
shaw, J., 1999. Adaptive management. In: Johnson, N.C., Malk,
A.J., Sexton, W., Szaro, R. (Eds.), Ecological Stewardship: A
Common Reference for Ecosystem Management, vol. 3. Else-
vier Press, Amsterdam, Netherlands, pp. 505–533.
P.F. Hessburg et al. / Forest Ecology and Management 211 (2005) 117–139 137
Boyd, R., 1999. Strategies of Indian burning in the Willamette
valley. In: Boyd, R. (Ed.), Indians, Fire, and the Land in the
Pacific Northwest. Oregon State University Press, Corvallis, OR,
pp. 94–138.
Camp, A.E., Oliver, C.D., Hessburg, P.F., Everett, R.L., 1997.
Predicting late-successional fire refugia from physiography
and topography. Forest Ecol. Manage. 95, 63–77.
Cissel, J.H., Swanson, F.J., Weisberg, P.J., 1999. Landscape man-
agement using historical fire regimes. Ecol. Appl. 9, 1217–1231.
Cooper, C.F, 1961a. Pattern in ponderosa pine forests. Ecology 42,
493–499.
Cooper, C.F., 1961b. The ecology of fire. Sci. Am. 204, 150–160.
Daubenmire, R., Daubenmire, J., 1968. Forest vegetation of eastern
Washington and northern Idaho. In: Technical Bulletin, vol. 60,
Washington State University Agricultural Experiment Station,
Pullman, WA.
Everett, R.L., Hessburg, P.F., Jensen, M.E., Bormann, B., 1994.
Executive Summary, vol. I. Eastside Forest Health Assessment.
General Technical Report PNW-GTR-317. U.S. Department of
Agriculture, Forest Service, Pacific Northwest Research Station,
Portland, OR, 61 p.
Finney, M.A., 2001. Design of regular landscape fuel treatment
patterns for modifying fire growth and behavior. Forest Sci. 47,
219–228.
Franklin, J.F., 1993. The fundamentals of ecosystem management
with applications in the Pacific Northwest. In: Aplet, G.H.,
Johnson, N., Olson, J.T., Sample, V.A. (Eds.), Defining Sustain-
able Forestry. Island Press, Washington, DC, pp. 127–144.
Franklin, J.F., Cromack, Jr., K., Denison, W.C., McKee, A., Maser,
C., Sedell, J., Swanson, F., Juday, G., 1981. Ecological char-
acteristics of old-growth Douglas-fir forests. General Technical
Report PNW-GTR-118. U.S. Department of Agriculture, For-
est Service, Pacific Northwest Research Station, Portland, OR,
48 p.
Franklin, J.F., Agee, J.K., 2003. Forging a science-based national
forest fire policy. Issues Sci. Technol. 20, 59–66.
Franklin, J.F., Dyrness, C.T., 1988. Natural Vegetation of Oregon
and Washington. Oregon State University Press, Corvallis, OR,
452 p..
Fule, P.Z., Covington, W.W., Moore, M.M., 1997. Determining
reference conditions for ecosystem management of southwes-
tern ponderosa pine. Ecol. Appl. 7, 895–908.
Gast, W.R., Scott, D.W., Schmitt, C.L., et al., 1991. Blue Mountains
forest health report: new perspectives in forest health. U.S.
Department of Agriculture, Forest Service, Pacific Northwest
Region, Malheur, Umatilla, and Wallowa-Whitman National
Forests Portland, OR.
Graham, R.T., Harvey, A.E., Jain, T.B., Tonn, J.R., 1999. The effects
of thinning and similar stand treatments on fire behavior in
Western forests. General Technical Report PNW-GTR-463. U.S.
Department of Agriculture, Forest Service, Pacific Northwest
Research Station, Portland, OR, 27 p.
Graham, R.T., McCaffrey, S., Jain, T.B. 2004. Science basis for
changing forest structure to modify wildfire behavior and sever-
ity. General Technical Report RMRS-GTR-120. U.S. Depart-
ment of Agriculture, Forest Service, Rocky Mountain Research
Station, Fort Collins, CO, 43 p.
Gruell, G.E., Schmidt, W.C., Arno, S.F., et al., 1982. Seventy years
of vegetative change in a managed ponderosa pine forest in
western Montana: implications for resource management. Gen-
eral Technical Report INT-130. U.S. Department of Agriculture,
Forest Service, Intermountain Forest and Range Experiment
Station, Ogden, UT, 42 p.
Hall, F.C., 1976. Fire and vegetation in the Blue Mountains:
implications for land managers. In: Proceedings of the 15th
Annual Tall Timbers Fire Ecology Conference, vol. 15, 16–17
October 1974, Tall Timbers Research Station, Portland, OR/
Tallahassee, FL, pp. 155–170.
Harrod, R.J., Bradner, H.M., Hartl, W.E., 1999. Historical stand
reconstruction in ponderosa pine forests to guide silvicultural
prescriptions. Forest Ecol. Manage. 114, 433–446.
Hann W.J., Jones, J.L., Karl, M.G., Hessburg, P.F., Keane, R.E.,
Long, D.G., Menakis, J.P., McNicoll, C.H., Leonard, S.G.,
Gravenmier, R.A., Smith, B.G., 1997. An assessment of land-
scape dynamics of the Basin. In: Quigley, A. (Ed.), An Assess-
ment of Ecosystem Components in the Interior Columbia Basin
and Portions of the Klamath and Great Basins. General Tech-
nical Report PNW-GTR-405. U.S. Department of Agriculture,
Forest Service, Pacific Northwest Research Station, Portland,
OR.
Hessburg, P.F., Smith, B.G., Kreiter, S.G., Miller, C.A., Salter, R.B.,
McNicholl, C.H., Hann, W.J., 1999a. Historical and current
forest and range landscapes in the Interior Columbia River Basin
and portions of the Klamath and Great Basins. Part 1: linking
vegetation patterns and landscape vulnerability to potential
insect and pathogen disturbances. General Technical Report
PNW-GTR-458. U.S. Department of Agriculture Forest Service,
Pacific Northwest Research Station, Portland, OR, 357 p.
Hessburg, P.F., Smith, B.G., Salter, R.B., 1999b. Detecting change in
forest spatial patterns from reference conditions. Ecol. Appl. 9,
1232–1252.
Hessburg, P.F., Smith, B.G., Salter, R.B., 1999c. Using estimates of
natural variation to detect ecologically important change in
forest spatial patterns: a case study, Cascade Range, Eastern
Washington. Research Paper PNW-RP-514. U.S. Department of
Agriculture Forest Service, Pacific Northwest Research Station,
Portland, OR, 65 p.
Hessburg, P.F., Smith, B.G., Miller, C.A., Kreiter, S.G., Salter, R.B.,
1999d. Modeling change in potential landscape vulnerability to
forest insect and pathogen disturbances: methods for forested
subwatersheds sampled in the mid-scale Interior Columbia River
Basin assessment. PNW-GTR-454. U.S. Department of Agri-
culture Forest Service, Pacific Northwest Research Station,
Portland, OR, 56 p.
Hessburg, P.F., Smith, B.G., Salter, R.B., Ottmar, R.D., Alvarado, E.,
2000. Recent changes (1930s–1990s) in spatial patterns of
interior Northwest forests, USA. Forest Ecol. Manage. 136,
53–83.
Hessburg, P.F., Agee, J.K., 2003. An environmental narrative of
Inland Northwest US forests, 1800–2000. Forest Ecol. Manage.
178, 23–59, Fire Aquat. Ecosyst. (Special Feature).
Hessburg, P.F., Reynolds, K.M., Salter, R.B., Richmond, M.B.,
2004. Using a decision support system to estimate departures
of present forest landscape patterns from historical conditions:
P.F. Hessburg et al. / Forest Ecology and Management 211 (2005) 117–139138
an example from the Inland Northwest region of the United
States. In: Perera, A.H., Buse, L.J., Weber, M.G. (Eds.), Emu-
lating Natural Forest Landscape Disturbances: Concepts and
Applications. Columbia University Press, New York, NY, United
States, (Chapter 13), pp. 158–174.
Huff, M.H., Ottmar, R.D., Lehmkuhl, J.F., Hessburg, P.F., Everett,
R.L., Alvarado, E., Vihanek, R.H., 1995. Historical and current
forest landscapes of eastern Oregon and Washington. Part II:
potential fire behavior and smoke production. In: Hessburg, P.F.
(Ed.), Eastside Forest Ecosystem Health Assessment, vol. III.
Assessment General Technical Report PNW-GTR-355. USDA,
Forest Service, Pacific Northwest Research Station, Portland,
OR, 43 p.
Irwin, L.L., Cook, J.G., Riggs, R.A, Skovlin, J.M., 1994. Effects of
long-term grazing by big game and livestock in the blue
mountains forest ecosystems. In: Hessburg, P.F. (Ed.), Eastside
Forest Ecosystem Health Assessment, vol. III, Assessment
General Technical Report PNW-325. U.S. Department of Agri-
culture, Forest Service, Pacific Northwest Research Station,
Portland, OR, 49 p.
Kauffman, J.B., 1990. Ecological relationships of vegetation and
fire in Pacific Northwest forests. In: Walstad, J.D., Radosevich,
S.R., Sandberg, D.V. (Eds.), Natural and Prescribed Fire in
Pacific Northwest Forests. Oregon State University Press,
Corvallis, OR.
Keane, R.E., Parsons, R., Hessburg, P.F., 2002. Estimating historical
range and variation of landscape patch dynamics: limitations of
the simulation approach. Ecol. Model. 151 (1), 29–49.
Landres, P.B., Morgan, P., Swanson, F.J., 1999. Overview of the use
of natural variability concepts in managing ecological systems.
Ecol. Appl. 9 (4), 1179–1188.
Lehmkuhl, J.F., Hessburg, P.F., Everett, R.L., et al., 1994. Historical
and current forest landscapes of eastern Oregon and Washington.
Part I: vegetation pattern and insect and disease hazards. General
Technical Report PNW-GTR-328. U.S. Department of Agricul-
ture, Forest Service, Pacific Northwest Research Station, Port-
land, OR, 88 p.
Lee, D.C., Sedell, J.R., Rieman, B.E., Thurow, R.F., Williams, J.E.,
1997. Broad-scale assessment of aquatic species and habitats.
General Technical Report PNW-GTR-405. U.S. Department of
Agriculture, Forest Service, Pacific Northwest Research Station,
Portland, OR (Chapter 4).
Lewontin, R.C., 1966. Is nature probable or capricious? Bioscience
16, 25–27.
Lindenmayer, D.B., Franklin, J.F., 2003. Conserving Forest Biodi-
versity: A Comprehensive Multi-Scale Approach. Island Press,
Washington, DC.
McCool, S.F., Burchfield, J.A., Allen, S.D., 1997. Social assessment.
In: Quigley, A. (Ed.). An Assessment of Ecosystem Components
in the Interior Columbia Basin and Portions of the Klamath and
Great Basins. PNW-GTR-405. U.S. Department of Agriculture,
Forest Service, Pacific Northwest Research Station, Portland,
OR (Chapter 7).
Morris, W.G., 1934a. Forest fires in Oregon and Washington.
Oregon Hist. Q. 35, 313–339.
Morris, W.G., 1934b. Lightning storms and fires on the National
Forests of Oregon and Washington. Unnumbered publication.
U.S. Department of Agriculture, Forest Service, Pacific North-
west Forest and Range Experiment Station, Portland, OR, 27 p.
O’Laughlin, J., MacCracken, J.G., Adams, D.L., et al., 1993. Forest
health conditions in Idaho. Report 11. ID: University of Idaho,
Forest, Wildlife, and Range Policy Analysis Group, Moscow, p.
244.
Perry, D.A., Henjum, M.G., Karr, J.R., Bottom, D.L., Bednarz, J.C.,
Wright, J.C., Beckwitt, S.A., Beckwitt, E., 1994. Interim protec-
tion for late-successional forests, fisheries, and watersheds: a
summary of the report of the eastside forests scientific society
panel. In: Everett, R.L., Baumgartner, D.M. (Eds.), Proceedings
of the Symposium on Ecosystem Management in Western
Interior Forests. pp. 103–114.
Pyne, S.J., 1982. Fire in America: A Cultural History of Wildland
and Rural Fire. Princeton University Press, Princeton, NJ.
Quigley, T.M., Haynes, R.W., Graham, R.T. (Eds.), 1996. Integrated
Scientific Assessment for Ecosystem Management in the Interior
Columbia Basin and Portions of the Klamath and Great Basins.
U.S. Department of Agriculture, Forest Service, Pacific North-
west Research Station, Portland, OR, PNW-GTR-382.
Reckhow, K.H., 1994. Importance of scientific uncertainty in deci-
sion making. Environ. Manage. 18, 161–166.
Reynolds, K.M., Hessburg, P.F., 2004. Decision support for inte-
grated landscape evaluation and restoration planning. Forest
Ecol. Manage. (invited feature).
Rieman, B.E., Hessburg, P.F., Lee, D.C., et al., 2000. Toward
an integrated classification of ecosystems: defining opportuni-
ties for managing fish and forests. Environ. Manage. 25, 425–
444.
Riggs, R.A., Tiedemann, A.R., Cook, J.G., Ballard, T.M., Edgerton,
P.J., Vavra, M., Krueger, W.C., Hall, F.C., Bryant, L.D., Irwin,
L.L., Delcurto, T., 2000. Modification of mixed-conifer forests
by ruminant herbivores in the Blue Mountains ecological pro-
vince. Research Patent PNW-RP-527. U.S. Department of Agri-
culture, Forest Service, Pacific Northwest Research Station,
Portland, OR, 77 p.
Robbins, W.G., 1999. Landscape and environment: ecological
change in the Intermontane Northwest. In: Boyd, R. (Ed.),
Indians, Fire, and the Land in the Pacific Northwest. Oregon
State University Press, Corvallis, OR, pp. 219–237.
Schmidt, W.C., Wakimoto, R.H., 1988. Cultural practices that can
reduce fire hazards to homes in the Interior West. In: Fischer,
W.C., Arno, S.F. (Eds.), Protecting People and Homes from
Wildfire in the Interior West. Proceedings from the Symposium
and Workshop. General Technical Report INT-GTR-251. U.S.
Department of Agriculture, Forest Service, Intermountain
Research Station, Ogden, UT, pp. 131–141.
Sierra Nevada Ecosystem Project (SNEP), 1996. Status of the Sierra
Nevada, vol. II. Assessments and Scientific Basis for Manage-
ment Options. Final Report to Congress. Wildland Resources
Center Report No. 37, University of California, Davis, CA, pp.
1–1528.
Soeriaatmadja, R.E., 1966. Fire History of the Ponderosa Pine
Forests of the Warm Springs Indian. .
Swetnam, T.W., Allen, C.D., Betancourt, J.L., 1999. Applied histor-
ical ecology: using the past to manage for the future. Ecol. Appl.
9 (4), 1189–1206.
P.F. Hessburg et al. / Forest Ecology and Management 211 (2005) 117–139 139
Swetnam, T.W., Lynch, A.M., 1993. Multi-century, regional-scale
patterns of western spruce budworm outbreaks. Ecol. Monogr.
63 (4), 399–424.
Thomas, J.W., Huke, S., 1996. The forest service approach to
healthy ecosystems. J. Forestry 94 (8), 14–18.
Thompson, I.D., Harestad, A.S., 2004. The ecological and genetic
basis for emulating natural disturbance in forest management.
In: Perera, A.H., Buse, L.J., Weber, M.G. (Eds.), Emulating
Natural Forest Landscape Disturbances: Concepts and Applica-
tions. Columbia University Press, New York, NY, United States,
(Chapter 3), pp. 29–42.
Turner, M.C., Romme, W.H., 1994. Landscape dynamics in crown
fire systems. Landscape Ecol. 9, 59–77.
USDA, 1993. A first approximation of ecosystem health. National
Forest System Lands, U.S. Department of Agriculture, Forest
Service, Pacific Northwest Region, Portland, OR, 109 p.
U.S. Government, 2003. Healthy Forests Restoration Act of 2003,
108th Congress, 1st Session. House of Representatives Report
108–386, pp. 1–48.
Walters, C.J., Hollings, C.S., 1990. Large-scale management experi-
ments and learning by doing. Ecology 71 (6), 2060–2068.
Weaver, H., 1959. Ecological changes in the ponderosa pine forest of
the warm springs Indian reservation in Oregon. J. Forestry 57,
15–20.
Weaver, H., 1961. Ecological changes in the ponderosa pine fore-
sts of Cedar Valley in southern Washington. Ecology 42, 416–
420.
Whitlock, C., Knox, M.A., 2002. Prehistoric burning in the Pacific
Northwest. In: Vale, T.R. (Ed.), Fire, Native Peoples, and the
Natural Landscape. Island Press, Washington, DC.
Wissmar, R.C., Smith, J.E., McIntosh, B.A., et al., 1994a. A history
of resource use and disturbance in riverine basins of eastern
Oregon and Washington (early 1800s–1990s). Northwest Sci.
68, 1–35.
Wu, J., Loucks, O.L., 1995. From balance of nature to hierarchical
patch dynamics: a paradigm shift in ecology. Q. Rev. Biol. 70,
439–466.
top related